基于改进的L_1范数非因果滤波器多次波自适应减方法  

An improved L_1-norm adaptive multiple subtraction method based on non-causal matching filter

作　　者：熊繁升 李荣贤 高孝巧[3] 

机构地区：[1]中国地质大学(北京)工程技术学院,北京100083 [2]云南省国土建设工程总公司,昆明650041 [3]中国地质大学(北京)地球科学与资源学院,北京100083

出　　处：《物探化探计算技术》2015年第2期215-223,共9页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：国家潜在油气资源产学研用合作创新项目(OSR-04-07);国家级大学生创新创业训练计划项目(201211415023)

摘　　要：多次波压制是地震勘探中的重要课题。基于匹配滤波器设计的多次波自适应减方法,对L1范数多次波自适应减方法进行改进,增加了一个门限值参数。同时针对常规自适应减方法中采用的因果滤波器只能沿时间轴正向进行多次波匹配的不足,采用了非因果滤波器进行多次波自适应减,增加了算子零时间的位置参数。通过引入新参数增强方法的适应性,改进方法提高了计算速度。通过对所设计的1D数据模型和合成地震记录的处理表明,在预测的多次波与原始记录中的多次波存在振幅和到达时间差别的情况下,非因果滤波器能给出较理想的结果;同时改进的自适应减方法可以减少计算时间。Multiple attenuation is an important issue in seismic exploration field.Adaptive multiple subtraction method cho-sen in this paper are based on matching filter,and then improved the L1 - norm adaptive emultiple subtraction method by adding a threshold parameter.While conventional causal matching filter for adaptive multiple subtraction can only be used in matching the multiple waves forward along the timeline.For this problem,we designed a non-causal matching filter to multi-ple subtractions,a parameter which shows the zero-time position was added.By introducing new parameters,the scope of the method was expanded;and the speed was faster after improving the method.After processing the 1D data model and synthetic seismogram record,we knew that while there are amplitude and arrival time differences between prediction multiple traces and the original multiple waves,non- causal matching filter could give a good result,and by improving the adaptive subtraction method could reduce the computation time.

关 键 词：多次波 自适应减 非因果滤波器 改进的L1范数 

分 类 号：P631.443[天文地球—地质矿产勘探]